I shot 60i for the reasons specified; a lot of people who use the XL2 favor the 30p and I just wanted to get that out of the way.

The video length is not certain yet; I still have 2 full days of shooting left, but I'm estimating time-wise it should be between 60-90 minutes. A DVD permits 4.3 gigs of data correct? Does linear time make a difference as well?

As for the border, indeed I have seen pro productions with these: Slim in 6 as an example has an graphical overlay border as do a few infomercial DVD exercise routines I've seen. I can most assuredly live without it, however I thought it might add some production value if done right. See image:

http://img404.imageshack.us/img404/4397/slim6.jpg

The longer the video, the lower the bitrate. MPEG2 compression is more of an art than a science. The ultimate image quality depends on many factors, bitrate being one of them. You should easily be able to get 60-90 minutes on a DVD with excellent quality. Use the bitrate calculator program available from my web site.

http://www.johncline.com/bitcalc110.zip

The border that was in the television monitor of the JPG to which you linked seemed to have some additional information in it, like time remaining for the excercise or the number of repetitions. This is a perfectly valid use of a border. It sounded to me like you were only trying to compensate for overscan and since you can't possibly know the worst case overscan on someone's TV, you can't really know how big to make the border to ensure that the full image displays on each and every TV out there.

The safe action and safe title area markers in Vegas are quite conservative, so if you really want to use the border then just make it fit in the outer "safe action" marker and make sure any text resides within the inner "safe title" marker. This should cover 99% of the TVs out there. The values are a percentage of the total screen dimensions. By default, Vegas considers safe action to be 10% in from the edge in both vertical and horizontal dimension and safe title to be 20% in from the edge. There is no official standard percentage, personally I changed the Vegas defaults to 5% and 10%. The BBC actually suggests 3.5% and 5%.

You've really opened up a can of worms for yourself...

If you planning on using a border, you can go two ways. You just the border as an image overlay which will cover up some of the footage that was shot.

Or you do a picture in picture where you have the fitness footage over the border image and size the frame of the fitness footage to fit your border. This will retain most of the edges of your video footage at the expense of size within the frame.

It looks like you could easily recreate the effect of the screenshot you posted using the second method above.

Gents,

Thanks for the info so far!

JC – The bitrate calculator you linked is a tremendously valuable tool indeed; thanks for that!
Respective to the border: As my video is indeed a fitness/workout production I had intended to utilize much of the space for data, i.e. reps, timers, logo, etc. As well, the cyclorama set we are using has imperfections abound and I wouldn’t mind having some elbow room to clip some undesirable peripheral portions of the clips if needed. Plus, the exercise product is demonstrated by one person at a time in the video; the wide 16:9 shots innately include a lot of superfluous peripheral oblivion in head-on shots which I would not mind clipping a bit.
My concern and reason for my question here is related to ample indent and scaling so as not to be too stingy or generous with said border. Additionally, I am concerned about image degradation should I need to shrink the video slightly to accommodate the nebulously defined border. Will scaling down the clips degrade the image quality?

All said, I am not absolutely committed to the border; I can just as easily float the same logo, timers, data, etc. atop the full screen image if absolutely need be.

Here’s my situation in a few paragraphs: I’m no video novice; I’ve been shooting intermittently since 2001; video is not my daily “bread and butter” but I’ve produced a number of quality, professional works for numerous business clients over the past 9 years with few complaints. Virtually all of the work I’ve produced has been designated for PCs – either on interactive CD-ROM, web-sites or for digital signage screens. In fact I started a DS network 4 years ago and have since sold it for a healthy profit. I was also approached by a prominent digital signage solution provider whom admired my work; I thereafter produced both their training video and their short-form expo PR piece.

Coming from an ENG style “straight to PC” video background admittedly imposes some habits which I now realize could pose challenges on this DVD project. I don’t know whether you’ve dabbled in DS production; I can tell you it is a WYSIWYG format. When I design a DS ad on my PC, every pixel fills the screen onto which it is delivered – not a single pixel lost to overscan. The same can be said for CD-ROM productions.
I indeed may have shot some scenes a bit too tight due to my aforementioned “straight to PC” habits coupled with the fact that the XL2’s viewfinder clips somewhere between 5-10% of what is actually being shot.

This particular project is personal; I’ve developed a fitness product (this thing may well go to infomercial/retail) and I’m the entire crew. I’ve partnered with a friend who is a former Olympian and accomplished martial artist (he is the fitness guru) and he’s conservatively bankrolling the project. This is a classic grass roots venture, and due to budget constraints every facet of this project respective to production is on me exclusively: from product design, CAD drawings, prototyping, materials sourcing, web-site and graphical design, packaging, copywriting, marketing, scripting, casting, studio procurement, lighting, shooting, editing, etc. etc!!! As such, the pressure I am under to juggle countless tasks simultaneously added to the unfamiliarity of a studio shoot with no assistance whatsoever makes for imperfect video acquisition. The clips themselves are excellent quality; I took pains in developing robust presets and tweaking camera settings to get the vivid, ethereal look I wanted from the white cyclo set. Nice as the clips are, I admit, I shot a number of them a little too tight having overlooked the overscan issue, and having the “active graphical border” in mind.

I played the clips on my plasma TV directly from the camera and they looked just fine. I know this TV has some degree of overscan from experience. I hope the DVD playback will not differ from the camera’s playback. I am still on the fence about the border, but much as the “Slim in 6” video, I can easily produce something similar and even matte in some subtle Digital Juice atop the border to bump up the production – we’ll see.

I must ask you JC – you’d stated Vegas’ safe areas are “conservative” yet you went on to say that you’ve reduced your safe area values from 10/20% to 5/10%. That said, wouldn’t you consider Vegas’ safe areas to be too liberal?

As well, I’m still unclear about the best project settings especially progressive or interlaced; would you mind offering some advice? The current V5 “NTSC DVD Widescreen” template uses interlacing and no deinterlace method.

Thanks,

Mov

"Will scaling down the clips degrade the image quality?"

Scaling the clips down will not affect the quality other than making the image smaller. Scaling up in size is where the image quality can suffer. Of course, for proper scaling of interlaced video you need to make sure the "deinterlace method" is set under your project's properties. "Interpolate" would be the appropriate method since your video has a lot of motion in it. Also, when rendering, make sure the render quality is set to "Best.

"That said, wouldn’t you consider Vegas’ safe areas to be too liberal?"

I guess it's a matter of semantics, is the glass half empty vs. half full? Vegas' 10%/20% makes sure it covers the absolute worst case overscan situation. I think it goes a little too far.

In order to have the smoothest possible motion, I would render interlaced which will mean 59.94 individual images per second as opposed to 29.97 if you render progressive.

The bitrate calculator you got from my web site defaults to calculations in 1024 bits. Vegas deals with it in bits, so you need to get into the bitrate calculator's settings and check the "1 kilobit = 1000 bits" setting. If the calculator says 6400, then you would enter that as "6,400,000" in Vegas' MPEG2 encoder. One more thing in the custom settings of Vegas MPEG2 encoder, make sure the quality slider is set to "31", which is its maximum. If your video ends up being over about 72 minutes, then you will want to use 2-pass VBR instead of CBR in order to maximize quality.

Regarding the border, you have convinced me that in your case it is a good idea since you will be using it to present additional information. If this project was exclusively for viewing on a computer monitor then there would be no reason to compensate for overscan.

JC - Just wanted to sincerely thank you for all of the great advice – much appreciated amigo!!

I'm fiddling with the layout/bitrates now. While the active info border looks great on DVD / TV and really enhances the production, I noticed the video portion of the clip looks "dirty." I wonder if using the interpolation could have caused this. By "dirty" I refer to artifacting in skin tones and an overall mosquito net muddying of the image – not abysmal, but not nearly as clean as the footage out of the camera.

I picked up 10 DVDs to burn and try to determine the best combos to optimize the video for crispness. The border overlay looks pretty clear and so does an animated title segment I rendered as an uncompressed .avi – only the video itself looks sub-standard. I used a CBR of 6,000,000 expecting that to be the approximate range I’ll be in if the finished video exceeds 60 minutes.

Any ideas what may be going on?

Thanks again John, you’ve helped me greatly.

PS - Oddly enough the same clip looks fine when it plays on my PC - it only looks "dirty" on DVD/TV (video portion primarily). I used DVDA 2.0 to create the simple disc for testing purposes.

Best,
Mov

What did you do to produce your video? Have you added any color correction or effects? Did you render your footage before placing in on the Timeline. Do you know about working in 16-235 (tv colorspace) vs 0-256 (computer colorspace)?

No pre-rendering, only mild brightness, saturation and sharpening applied to the footage imported directly from the camera. I know a little about the different color spaces...what are you getting at respective to the issue I seem to be having?

It's probably the sharpening that's getting you in trouble.

So far not too good.

I've rendered several versions of the clip using a combination of different settings. It seems that rescaling the image does indeed soften it and sharpening is required - I can see the difference in the V5 preview window and in rendered clips. Clips rendered with interpolation and no sharpening look plain lousy when played on PC and on DVD. I even rendered without interpolation; while the image quality is greatly improved, the motion shows weird banding on the TV screen.

Something’s amiss here; I have good quality footage which looks VERY clean right out of the camera. After performing 3 very minor and typical adjustments in V5, the footage looks as if it were shot with a Handycam!! After designing and integrating the info border, I feel it adds a lot of production value and I really want to keep it. Either V5, the mpeg-2 codec or a combination of the two is compromising the image and I’m kind of freaked out at this moment.

I’ve seen footage from $500.00 consumer camcorders look vastly better than my professional footage rendered out of V5 and DVDA 2.0 scaled for the info border.

I wonder if rendering each segment from the V5 timeline as uncompressed .avi and using a better mpeg-2 encoder would help???

There has to be a solution for this problem!! Any help would be greatly appreciated.

Mov

Are you really still using Vegas 5 and DVD-A 2?

I used a CBR of 6,000,000 expecting that to be the approximate range I’ll be in if the finished video exceeds 60 minutes.

Movick, if the final length is under 70 min., use a CBR of 8,000,000.
For 90 min., use a VBR setting of 7,904,000 / 6,320,000 / 3,972,000
If any part of the video is of questionable quality, use the 2-pass option.
HTH.

Mike

PD - Yes I am; so what? Are the updates in the newer versions so mind blowing that it justifies the purchase price for someone who doesn't do all that much video production? I’m still shooting with a sparkling, low-mileage XL-2 (SD), and unless a newer version of Vegas will provide me with a better MPEG-2 codec, I don’t see the point.

If the current fitness product project goes well I’m going to get geared up for HD: new super fast PC or maybe even a Mac, a serious pro HD camera, etc. At that time I will either consider the latest version of Vegas or totally switch over to FCP as so many have suggested. For now V5 is familiar and other than the MPEG-2 quality which I understand is an entity unto itself, it works just fine. I remember upgrading in the past from earlier versions of Vegas and quite frankly the updates were not useful enough to justify many of the beta glitches with which I had to contend.

Everyone these days has a need to get the latest & greatest, often without adequate justification. In all honesty, if one is competent he/she can get the job done with the fundamental tools. As an example, the TV series “House” is shot exclusively on a Canon 5D MKII which is a DSRL with HD acquisition capabilities. No CineAlta cams on that set!

I am all ears…well in this case eyes…what am I missing in the newer versions of Vegas which will dramatically improve my work?

Mike - I have used the higher CBR and it doesn't make much difference unfortunately. Clearly the MPEG-2 encoding in V5 isn't that fantastic. From what I've read MPEG-2 encoding is an art; high-end hardware can run up to 40K. If all else fails I can probably purchase a decent piece of software or use a service to encode from .avi. I would rather not have to deal with that if I can remedy this in Vegas.

Thanks for the tips!

Mov

I have no experience with V5 (I started briefly with V7). I was just flagging it in case any "old timers" knew of any issues with that version.

In shrinking your video to accommodate your border you are effectively reducing its resolution. I wonder wether your problem is analogous to the problem of HD to SD conversion. Some people claim that it is necessary to first deinterlace to double rate (i.e. 60i to 60p) using a good deinterlacer (motion compensated?), then down rez and convert back to interlaced again. Did you say whether on not you get good results if you don't shrink your video?